They've got a hunch on you

Hunch is a fairly new player in the field of social media but was ranked as one of the world's most intriguing start-ups in Busines Week just last year. It's basically a recommendation engine which personalizes a taste profile of the user by asking questions. Then it gives recommendations based on the decision pattern of people similar to the user. Thanks to this matching taste graph logic, it can predict answer to user's problem. Or can it?

So you answer a bunch of questions and get a solution. Easy. The more you use it, the more accurate it is. The bigger and active the user community, the better it works. In addition to Teach Hunch About You (THAY) questions, you can educate the engine by fixing its default beliefs of you around the topic, e.g. sports. Also, the service pulls in data about articles, topics and people that you and your friends “like” on Facebook or follow on Twitter if user logs on through these profiles. 

User can suggest and produce topics. Co-creation and participation are brought forth, encouraged and emphasized on the site. Hunch has its own 'member programme' in which active users receive creds and banjos. Yes, banjos! It also provides heavy users with profiling padges such as Inventor, Early Bird, Curious and Good Citizen. So why should I participate? you may find yourself asking after 10 banjos. Everybody knows how empty it feels when nobody notes your updates, posts or ideas. But no worries as hunch also understands the importance of feedback. When users thumb up hunches and share views within topics, the initiator is informed as well. 

So how did hunch work for me? I find answering to random questions rather entertaining and addictive. The first session of 20 questions doesn't take too much time and the results are shocking enough to trigger my interest. It also is intriguing tool to reflect yourself to other users. Whilst you go about the never ending THAY questions, it's  fun to know how common your choice in fact was at the time of answering  For example, it's kind of nice to know that 51% of hunch users use the same smiley that I do and that only 39% of the respondents twitter. What does Who you follow say about You? asks hunch's Twitter Follower Stats application where you can compare answers to the ones of other twitterers'. For example, the followers of The Economist strive for power over riches, fame and security. The ones of Barack Obama, on the other hand, don't use the phrase LOL. 

Conveniently hunch doesn't allow any ads on its site but sponsored links are presented in some of the topics such as travelling locations. Ultimately hunch is a behavioral data gem as B2B service and especially to e-commerce. With behavioral data firms can find more relevant consumer groups. Hunch sells external companies taste graphs and other aggregated information with which companies can also define their recommendations on their own sites. This month TechCrunch announced hunch's new partnerships, but partnership co-operation is yet to be explained on the site.

But again the question on everybody's mind: when you proactively educate a engine about you, can it really recommend anything new? So far hunch's predictions have fallen to things-I-already-know category. Moreover, I find the topic descriptions rather dull - most of them copied from Wikipedia. IMHO an engine telling people what they want doesn't hook for long unless there is a twist. So hunch is dependent on the community not only volume but also interesting content wise. 

Hunch wants to give smart recommendations and ease the search process. It is true that search engine results usually include spam and finding pleasing information can take a lot of energy. But in the end, is hunch the new influencer and do the recommendations result in any actions? Let's stay tuned!

Check out useful articles here and here!